🎤Voice Biometrics Authentication

---

This API verifies a user by comparing both their voice print and spoken phrase,and is
 used each time the user attempts to log in.

Voice Biometrics Authentication

PUT PUT /login-app/

Query Parameters

Name	Type	Required	Description
collection_name	string	✅	Company name: (will be provided)
login_phrase	string	✅	Expected phrase the user must say

Request Body

Name	Type	Description
data	object	Object containing the parameters below: { "RecordData_app": "<base64-encoded WAV audio>" }

Status Codes

✅200 Successful

{
  "data": {
    "status": "Login successful",
    "thresholds": {
      "vector_similarity": "0.91",
      "phrase_similarity": "0.88"
    }
  },
  "errors": {
    "code": "",
    "message": ""
  }
}

Request Body (RecordData)

{
  "RecordData_app": "<base64-encoded WAV audio>"
}

Example of request data:

(login_data: RecordData_app, register_data: RecordData_app, login_phrase, collection_name)

❗Error Format

All API responses follow this format, even on failure:

{ "data": "", "errors": { "code": 702, "message": "Error during login: speaker embedding mismatch" } }

{
  "data": "",
  "errors": {
    "code": 702,
    "message": "Error during login: speaker embedding mismatch"
  }
}

🧾 Error Codes

Code	Identifier	Description
700	GENERAL_COLLECTION_ERROR	Failed to create or manage a Milvus collection
701	GENERAL_REGISTERING_ERROR	Failed to extract or store speaker data
702	GENERAL_LOGIN_ERROR	General failure during login
703	MISSING_LOGIN_INFORMATION	Login phrase or voice input is missing/invalid